Hey There: Please forgive the 'sketchyness' of this "train-of-thoughts" that may follow in my rush to reply to your idea, but this is the only time i have to respond and your post is of great interest to me.
Reading below and ignoring the tech specs for a moment, it seems that you are envisioning a machine that can ultimately become self-aware in behaving as a proxy for it's user. My wife and I were discussing something related yesterday. To wit: that in attempting to become better organized, it's often frustrating to be spending as much or more time managing the list of jobs/tasks/activities as actually doing them. Setting priorities on the items, granularizing them into sub-tasks, what-ever, focusing on secondary-levels of work tend to become more time-consuming and detracts from the pleasure of actually doing the real work: the J|T|A tends to be perceived more perfunctorily, something to check off the list, than something to enjoy doing. A similiar sentiment was expressed (by you?) at the last meet when i suggested the necessity of creating a specification (meta-data) around which to build up the koha deployment; the counter arg being "skip that and just enjoy the 'doing' of the application. I now understand the merits of that attitude as trying to define and understand and generally "know more" can become tiresome and a diversion from the doing. This is analogous to computing at one's workstation and either doing real work or constantly tweaking their env to (hopefully) become more efficient, more organized; as the workload and everything related to it grows, and grows, and grows... Like tagging, its' akin to adding 100's or 1000's of softlinks cross-ref'ing one's FSys; or some mechanism that bridges application-space with data-space as part of one's KBase (e.g. storing|organizing what i need to know about trac to keep it running AOT what trac actually contains) A lot of what you mention is about ergonomics, GUI-vs-CLI, button steps AOT an alias that pipes multiple funcs, touch-vs-click (or voice vs touch), the paper-clip as your friend:) But the efficacy of computing will be marked by the level to which we use computers. For those (non-tech) ppl who sit in front of a screen 9-5 then turn it off and go home, it may be that there are ways to increase productivity, which, IMO, is what this thread boils down to. For them, a computer is little more than an appliance. For myself, and I suspect for many on this list, my computer is an integral part of me. What it contains|stores and the applications that operate on that data are an extension of myself. It reflects and IS my work, my fun, my entertainment, my life. Sure, I get out now and then; but if I go camping say, the box holds the photos of my experience. If I volunteer at the soup-kitchen, my box holds the correspondences with the people standing next to me. So, my BOX, becomes an extension of my HEAD and my memories. It has become an integral part of my life (I dont know if that is a good thing, but it is what it is) It contains years worth of 'stuff' that marks where i've been and what i've done and who it's involved. So, I honestly doubt that the 'context-sensitive' paradyme you envision would work for me; but I would welcome anything that improves organization and promotes greater efficiency in getting work done. Till now, I've just viewed it as unavoidable, like having to take whatever time is required to create a spec or proper documentation; its secondary to the actual code (process), but just as important. Until the day that AI can understand my actions and augment my workflow and is equipped w/voice recog and speech synth, I suspect that any tool/device that employs what you are describing has the potential to suck up one's time (the tweak factor) as much as what's already in place. my .02, Rion On Sunday 28 December 2008, Alvin ONeal wrote: > I have a plan to create a different kind of computer at some point in the > future or at least patent some sketches and bullet points I have written > down in a notebook so that Microsoft and Apple can't come up with them and > patent them before I have the millions I need to really get things going (of > course, I'd have some sort of deal where true open-source developers could > use the ideas royalty free). > > The buzz-word to describe what I imagine (and this has shown up on the Gnome > 3 wiki dev notes) is context-sensitive "task-based computing". Gnome-Do, > Ubuntu's Netbook launcher, Picasa, Google Desktop, MarkLogic Server, the > iTouch, DropBox, SimplifyMedia, and Prizm all have their foot in the door on > this concept but none of them really do more than graze on individual pieces > of it. It has to be something so simple and easy that a geek would prefer it > over the commandline and actually be able to get things done with it faster. > > That's the real test. Any non-techie can fall for a flashy screen that tries > to guess at what you want to do, but to have something that even people that > know what they're doing prefer to use it - that's the Holy Grail in my mind. > I mean, I often open firefox from the terminal because going to the internet > menu takes to long, I don't want to go to the desktop area, and my mouse is > closer to the terminal window than the panel (though in the last few days > I've started the conversion process to Gnome-Do and I'm quite impressed - > it's context-sensitive in both directions, but has a few ugly limitations) > > And in the future I don't think that we'll be using mice and I don't think > that /home will be formatted with ext4, it will need a more database and > tag-system enabled/extended fs because using i_notify in the kernel to > notify 10 different applications every time that a file's meta-data changes > just isn't efficient... or maybe an indexing system that works somewhat like > an encrypted folder in that it could be applied to individual folders > ~/Pictures, ~/Music, ~/Documents, etc. > > Technology has advanced beyond the point that we should think of a computer > as a type a 'desktop' that follows the same conventions as a desk's desktop > that just has random crap strewn about it and is cleaned once every quarter. > What good does that do me? Then I just end up using 'locate' to find > something that is right in front of me because I don't remember putting it > there. And people won't organize a computer like a real desktop anyway > because they can't touch it and it's not 3d. It's too much time and effort > to go through the stuff. > > Case in point: Touch screens. Right now touch your LCD screen in different > places and imagine you are performing actions that way instead of a mouse. > Sucks, huh? Now tilt the head of your LCD at a 30 degree angle and try > again. Wow! But no one has invented this yet because no one has noticed that > an LCD screen isn't a TV with a bulky CRT that can't be angled and touched. > > Of course, I think that Linux is the only OS that could support a truly new > paradigm of computer because it's the only one modular enough to be able to > keep most existing applications as they are and switch out a few libraries > here and there to add the rudimentry functionality of what the PC > Environment of the future should be until it actually happens and new apps > are made specifically to take advantage of it. > > Anyway, if any of that seems to make sense and you happen to have a few > million to throw around or know someone who, let's talk sometime. I'd be > happy to drop out of school and become Bill Gates' anti-thesis. > > AJ ONeal >
signature.asc
Description: This is a digitally signed message part.
